Skip to main content

Tranche risk

Each vault splits into two tranches sharing one pool of strategies: Stable and Alpha. Every settlement, the waterfall decides who gets the profit and who eats the loss. When Alpha loses money. Alpha moves at a multiple of the vault, the inverse of its share of the vault. At a 20% Alpha share, Alpha moves 5x. Example: a 1,000,000 HONEY vault, 800,000 Stable / 200,000 Alpha. Strategies drop 25%, a loss of 250,000. Alpha absorbs the first 200,000 and goes to zero. The remaining 50,000 comes out of Stable, a 6.25% loss for Stable. There’s no clawback: a later profitable settlement pays Stable and Alpha as normal, with no memory of what Stable lost. Stable is not risk-free. Its target is a cap on the upside, not a guarantee. If profit falls short of the target, Stable takes whatever profit exists and the rest is forfeited, not carried to the next settlement. Stable earns nothing in a flat or losing period. And coverage - Alpha’s share of the vault - is observed in the app, not enforced on-chain: nothing stops it from drifting down, which raises Stable’s real exposure.

Smart contract risk

Berafarm’s contracts are new and not yet reviewed by a third-party auditor. → [Audits] Live vaults currently deploy into three external protocols: Kodiak (concentrated liquidity), Dolomite (lending), and Bend (lending, via Re7 and Clearstar). A bug or exploit in any of them is a loss to the vault, absorbed by Alpha first. More protocols may be added over time — check → [Contract addresses] for the live set.

Strategy and market risk

  • Impermanent loss on Kodiak’s concentrated LP sleeves when the paired assets move apart in price.
  • Depeg risk on any paired asset the vault holds (HONEY, USDe, stables).
  • Incentive decay : much of the yield is emissions-based and can shrink; Alpha feels this first, since Stable is paid before Alpha sees anything.
  • Liquidity depth : sleeve size isn’t bounded by pool depth, only by the vault’s TVL and per-user limits.

Withdrawal and liquidity risk

Deposits and withdrawals are instant, with no queue. That also means your withdrawal unwinds real positions live, at whatever the market gives you right then. You can receive less than your shares are worth. Shares burn for a claim at the current share price; you’re paid what the unwind actually realizes, capped at that claim. A large withdrawal against thin liquidity, or a paused sleeve being skipped in the unwind, both mean you realize less. A global pause (→ [Admin controls]) blocks withdrawals entirely. Exit fees apply only inside a time window after deposit; current windows are on → [Admin controls].

Oracle and pricing risk

Share price is book value divided by share supply, and book value comes from the strategies’ reported value. LP sleeves are valued using oracle prices, not the pool’s live spot price deliberately, to resist same-block price manipulation. That makes the oracle itself a dependency: a stale or invalid price feed blocks deposits and withdrawals until it recovers; a live-but-wrong price feed can mis-value the vault while it’s within its staleness window. Feed configuration can change instantly, with no delay.

Governance and parameter risk

Two things worth flagging: the Stable target rate has no upper bound, set too high. It starves Alpha or overflows and bricks the vault. And slippage tolerances can legally be set to 100%, accepting near-zero output on a swap.

Operational and key-person risk

Berafarm is actively managed. We choose the sleeves, set the weights, and decide when to rotate, team judgment is a risk factor a passive vault doesn’t carry. Rotation happens in one transaction at whatever slippage is configured. Upgrade authority sits behind a Safe multisig with no timelock; → [Admin controls] covers key holders. What the contract enforces (the ceilings and delays above) is different from what the team commits to (announcing rotations, using pause switches only in emergencies). The second is policy, and policy can change.

Verify it yourself